Secondary analysis of the DOSS study data: reducing noise
Research type
Research Study
Full title
Secondary analysis of the Delirium Observational Screening Scale (DOSS) study data: identifying data patterns indicating onset of delirium in care home residents with, and without dementia using the QuantiCode algorithm.
IRAS ID
230105
Contact name
Elizabeth Teale
Contact email
Sponsor organisation
Bradford Teaching Hospitals NHS Foundation Trust
Duration of Study in the UK
0 years, 6 months, 0 days
Research summary
The DOSS study (completed in November 2016) sought to determine whether a checklist of observed behaviours (the Delirium Observation Screening Scale) could detect delirium in residents of care homes. We propose a secondary analysis of anonymized data collected during the DOSS study.
Delirium is a common and deeply distressing condition characterised by sudden onset of confusion and behavioural disturbance. It is associated with poor outcomes (death, new dementia, dependency) and usually occurs when people are unwell (e.g. an infection or constipation).
Delirium is particularly common in older people with dementia and although there are shared features, the conditions are not the same. Detecting delirium in people with dementia is therefore very challenging. Failure to recognize delirium is associated with poorer outcomes; if identified, some of the contributing factors may be reversible.
During the DOSS project, care home staff completed daily DOSS assessments in 216 residents over several months. These repeated daily measurements are known as longitudinal data. The change in a DOSS score for an individual with dementia over time is likely to give more information than the average change in DOSS scores for the study population as a whole, as participants with dementia may have behavioural disturbances at their normal baseline. We aim to exploit changes in DOSS scores to detect delirium in people with and without dementia.
Working with mathematicians at the University of Leeds we plan to re-analyse the existing DOSS study data. As part of an existing research programme, this team have developed a new algorithm able to reduce the 'noise' caused by fluctuations in repeated measurements. This has wide applications in health and social care and commercially. We aim to use the DOSS data as a case-study to test this algorithm, at the same time allowing us to better answer our original research questions.REC name
North West - Haydock Research Ethics Committee
REC reference
17/NW/0416
Date of REC Opinion
3 Jul 2017
REC opinion
Favourable Opinion